Source: Three rounds of BIDS survey.

One may note that there are grey areas in above groupings. For example, some of the recent drop-outs may join the program in future, and therefore, belong to the category of occasional participants. Similarly, some of the recent participants may drop-out in future and join the group of either complete drop-outs or occasional participants. In spite of this problem, it enables a clear distinction between three important groups: regular participants, complete drop-outs and occasional participants, who together constitute 78 percent of the sample households in program villages. Since the “complete drop-outs” may rejoin the MFI net, one may aggregate “complete drop-outs”, “recent drop-outs” and “occasional participants”. We do this and identify the set as occasional participants. In most of the analyses to be presented in the subsequent chapters, we make use of three sample groups: regular participants, occasional participants and non-participants.

Finally, a pragmatic side of data management in longitudinal surveys needs to be noted. Households, which are the sampling units, are not stable either. With demographic transition, they split into two or more units. Moreover, some of the households covered during the first round of survey may not be available in the locality during subsequent rounds [3] . Anticipating these problems, the sample size was purposively chosen to be larger than that required by the terms of reference. Of the 3026 households, 112 (3.7 percent) split into two or more units (Table 2.3), while records for all three rounds were available for 2903 of the original units (including those which split, but information on all split units were available). In case of the split units, where applicable, we have aggregated the information on the split units to arrive at information on the original unit. Thus, most exercises in this report are based on the 2903 households. Village-specific information on the numbers of included households are presented in Table A.2.4 in the annexure at the end of this chapter.

Table 2.3
Distribution of Survey Households in terms of Stability of the Units and Availability During the Survey Rounds
 
Household stability
Recorded in all rounds
Missing in one or more rounds
Row total
Split into 2 or more units
76
36
112
Remained a single unit
2827
87
2914
Total number of households
2903
123
3026

Note: Details, by study villages, are provided in Table 2.A.4.    

Annexure to Chapter 2

Statistical Tables 
Table 2.A.1
Number of Surveyed Households in the Program Villages, by Participation Status
Region
Regular participants
Complete drop-out
Recent drop-outs
New participants
Non-participants
Occasional participants
Number of households
Panchagarh 92 3 11 3 56 28 193
Kurigram 90 8 13 5 52 34 202
Sahajadpur 61 4 25   57 50 197
Meherpur 60 11 27 1 48 51 198
Madaripur 101 3 9 8 53 27 201
Agailjhara 87 9 8 3 56 32 195
Tangail 104 2 9 4 46 30 195
Netrokona 75 2 15 4 52 45 193
Satkhira 89 12 12 5 46 29 193
Munshiganj 60 3 21 3 56 49 192
Chittagong 58 8 21 6 81 31 205
Feni 65 3 16 5 69 37 195
Bogra 119 8 20 4 71 44 266
Number of households 1061 76 207 51 743 487 2625
Percentages 40.42 2.90 7.89 1.94 28.30 18.55 100

 

Note: 1=regular (participants in all three rounds); 2= complete drop out (participated before the first survey and non-participant in all three rounds); 3 = recent drop-outs (participant during the first survey and non-participants

in the subsequent two rounds); 4 = new participants (previously non-participant and participant in the last round only); 5=non-participant all through; 6 = occasional participants.

Table 2.A.2
Number of Surveyed Households in the Control Villages, by Participation Status

Region

Regular participants

New participants

Non-participants

Occasional participants

Number of households

Sahajadpur

47

4

51

Tangail

2

2

20

25

49

Netrokona

2

47

1

50

Satkhira

11

31

9

51

Munshiganj

2

46

48

Chittagong

3

22

4

29

Number of hhs

5

17

213

43

278

Percentages

1.80

6.12

76.62

15.47

100.00


Table 2.A.3
Factors underlying Missing Households

Reasons for MissingPercent
Joined son’s family after death of husband1.1
Temporarily migrated out to work in urban area11.0
Family migrated to the work place of household head*7.7
Absence of members; left due to family strife8.8
Reunion of households 4.4
Left for being unable to repay debt*12.1
Shifted to other village in fear of river erosion11.0
Due to lack of homestead land, migrated to other villages2.2
Migrated to other countries*8.8
Returned to own village after residing in wife’s village11.0
Moved to government housing project area4.4
Not known*17.6

Note: Those with * are likely to have left the rural area. They sum up to 46.2%.

Source: BIDS Surveys in 1999 and 2000.

Table 2.A.4  
Distribution of sample Households in terms of Availability of Data in 3 Rounds and Household Splits

Region
Village
Undivided, available in all rounds
Undivided, dropped in 2nd round
Undivided; dropped in third round
Divided; available in all rounds
Undivided; dropped in 2nd and covered in 3rd round
Divided; missing in one of the later rounds
Sample size in 1st round

Panchagar

(Anuvab)

Baguladangi
30
1
1
 
   
2
34
Balabhir
33
   
   
1

   
 
34
Piramood
31
 
1
2 
 
  
34
Daksa para
30
 
 
1
   
3
34
Sarkar par
31
 
 1
2
 
1
35
Hasanpur
30
1
 
2
 
  
33

 

185
2
3
8
6
204

Kurigram

(Solidarity)

Pathak Para
32
 
 
3
 
 
35
Sakoa
31
1
 
 
 
 
34
Asharu
34
 
2
 
 
 
34
Joykumar
32
 
 
2
 
 
34
Paramanik Tari
34
 
 
 
 
 
34
Madajal Fara
34
 
 
 
 
 
34
 
197
1
2
5
 
 
205 
Shahazadpur
Ratankandi
30
 
 
2
 
 
32
Jugnidaha
29
 
 
1 
 
4
34
Prannathpur
35
 
 
 
 
1
36
Daya Islampur
31
 
 
2
 
1
34
(PPD)
Choto Shelachapri
31
 
1 
2
 
 
34
Nukali
33
 
 
1
 
   
34
Nandalalpur
23
 
1
2 
 
 
26
Boro Shelachapri
25
 
  
1
 
 
26

Meherpur

(PRP)

Kamdevpur
32
 
1
2 
 
1
 
Dariapur
33
 
 
 
  
   
 
Khandakar Para
34
   
 
 
 
 
 
Terogharia
29
 
4 
 
 
1
 
Radhakanta
pur
31
 
1
2
 
 
 
Kuhksa
34
 
   
1
 
 
 
   
193
  
6
5
  
2
206
 
  
237
2
11
6
256

Madaripur

(GUP)

Mohhah Kandi
33
 
 
1